Background

The state of Oregon is home to 50,500 students characterized as English Learners across its ~200 districts. These students represent close to 9% of reported students in Oregon schools, across all grades.

Initial Versions of the graph chart EL populations compared to total student population of each district.

Our second visualization plots raw counts of students across Oregon, showing areas of greater density. These counts are not intuitive, and make more sense when displayed relative to the overall number of students.

Understanding that the relative need for EL student support varies dramatically across Oregon allows for allocating resources to better serve as many students as possible, while acknowledging what more isolated EL students may face.

Background

Research Area: Examining concentration of dual language immersion programs in Oregon. Dual language immersion programs aim to promote high academic achievement, full biliteracy, and develop social consciousness by grouping EL and monolingual students into one classroom while splitting instruction in English and a partner language. Evidence from rigorous studies finds that TWDLI programs has a positive impact on the academic achievement of EL classified students (Steele, et al., 2017) and meta-analysis that examine the effects of bilingual education also finds a positive relationship (Rolstad et. al., 2018).

When I started this work, my intention was to examine if there are any patterns in the concentration of district with DLI programs by linguistic and racial diversity. More specifically, I wanted to know if districts with high EL and racial diversity were benefiting from these programs. My first graph plotted concentration of EL against pct white in the y-axis. Knowing that some districts might have low EL rates but a lot of EL students, I also included the raw number of students in the graph by tying size of points to raw count of EL students.

I was also thinking of dividing districts into quadrants and having four typologies of districts. However, I had to drop that concept because there weren’t really four types of districts, the relationship between x and y variables were fairly linear.

I ended up with two final graphs that incorporated many improvements relative to the first iteration of the graph. First, I changed the colors and adjusted the sizing scale so that they are more prominent and we can see differences. I also used ggrepel to highlight specific districts. Additionally, I changed the y-axis to center the message more on linguistically/racially diverse students and remove focus from white students. Lastly, I added a descriptive title that provided a key takeaway.

The first final graph highlighted where the DLI districts are located and emphasizes that racially & linguistically diverse districts have the largest programs. For example, Portland, Beaverton, Hillsboro, Salem-Keizer, and Woodburn all have large DLI programs and serve a high proportion of EL students.

The second graph kept many of these changes but communicated an additional finding which is that there are at least seven districts in Oregon that serve a high number of EL students but don’t offer DLI programs. These districts serve more than 500 current EL students and have high linguistic and racial diversity. Research indicates that many of these students could benefit from programs such as DLI.